Long-term curve — EIOPA methodology to 100 years
The daily LW-NSS Polish sovereign curve extrapolated to 100 years with the Smith-Wilson method under the EIOPA Solvency II parameter set for the zloty: Ultimate Forward Rate 3.30%, Last Liquid Point 10 years, convergence of the forward intensity to the UFR within 1 bp at 60 years. EIOPA itself derives the PLN risk-free curve from government bonds, the same market this site's curve fits.
How to read this page (methodology and glossary)
EIOPA's Solvency II risk-free term structures extrapolate beyond the Last Liquid Point with the Smith-Wilson method: liquid market rates up to the LLP are fitted exactly, and the forward intensity converges to the Ultimate Forward Rate, a long-term macro equilibrium equal to an expected real rate (common to all currencies, from a 1961-onward average) plus the currency's expected inflation (the central bank target mapped into buckets). For the zloty the applicable UFR is 3.30% for 2026 and 2027, the LLP is 10 years, and the convergence point is max(LLP+40, 60) = 60 years with a 1 bp tolerance on the forward intensity; the speed parameter α is the lowest value (≥ 0.05) meeting that tolerance. This page applies exactly that machinery to the LW-NSS curve: zero rates at tenors 1 to 10 years are converted to annual compounding, fitted exactly, and extrapolated.
What this is not: the official EIOPA PLN risk-free structure. The credit-risk adjustment from EIOPA's latest monthly file (currently 10 bp for the zloty) is deducted from the inputs as the Technical Documentation prescribes, and the official end-of-month EIOPA curve is overlaid for comparison, but EIOPA's bond inputs and monthly snapshot differ from the daily BondSpot fixings behind the LW-NSS fit. Treat the blue curve as the EIOPA methodology applied to this site's curve on today's data; the salmon overlay is the official structure at its own, older, month-end date.
Glossary
| UFR | Ultimate Forward Rate: expected real rate plus expected inflation, reviewed annually with a 15 bp annual change cap. PLN 3.30% (inflation bucket 2%, as for the euro; the NBP target is 2.5% ± 1 pp). |
| LLP | Last Liquid Point: the longest maturity whose market EIOPA's DLT assessment deems deep, liquid and transparent. PLN: 10 years, from government bonds. |
| Smith-Wilson | Exact-fit interpolation and extrapolation on discount factors, with a kernel controlled by α and an asymptotic forward intensity of ln(1+UFR). |
| convergence | The forward intensity must be within 1 bp of the UFR at year 60; α is the smallest value (≥ 0.05) achieving this. |
| forward intensity | The instantaneous forward rate implied by the extrapolated discount curve, shown annualised. |
Zero curve to 100 years
All coloured lines except the salmon one are this site's curve, drawn from one month of maturity, where EIOPA's official structure only begins at one year. Green: the LW-NSS zeros repriced exactly by Smith-Wilson up to the 10y LLP and extrapolated beyond it under the EIOPA parameters. Blue: the Smith-Wilson extrapolation beyond the LLP under the EIOPA parameters. Dotted: the same NSS parameters evaluated past 10y on their own functional form, with no regulatory anchor. Light-blue dash-dot: this site's curve on the EIOPA snapshot date, run through the identical CRA + Smith-Wilson recipe, so its gap to the salmon line isolates the bond-basket difference. Salmon: the official EIOPA PLN structure (no VA) from the latest monthly file. Annual compounding, percent.
Forward intensity and the glide to the UFR
The instantaneous forward rate of the extrapolated curve. By construction it reaches the UFR within 1 bp at year 60 and stays there.
Ours vs EIOPA's: a 48-month decomposition (the difference lives in the first year)
The LW-NSS curve is glued to the money market: its short end is anchored to the NBP reference rate and fitted through the full BondSpot bond panel. EIOPA's official PLN structure effectively lets the first twelve months of the curve go unobserved, since its shortest input is the 1y benchmark bond, an instrument that persistently trades rich on switch-auction buybacks and bank collateral demand. Across the 48 month-ends from August 2022 to July 2026 (both curves post-CRA at 10 bp, both annual compounding), that difference in short-end treatment costs on average +24 bp at the 1y point (0 to +71 bp, widest during the 2025 short-bond squeeze). Beyond the first year the two curves essentially agree: the average gap is +2 bp at 2y, 0 bp at 5y and +4 bp at 10y (standard deviations 2 to 6 bp), so once the first-year treatment is accounted for there is no residual basis between the LW-NSS fit and EIOPA's inputs worth naming. Substituting EIOPA's 1y rate for our first year while keeping our forwards confirms the mechanism: it moves the 2y point by about 17 bp and the effect fades by construction along the curve. Neither treatment is an error: EIOPA prioritises quote availability at benchmark tenors, the LW-NSS fit prioritises consistency with the policy rate and the full traded cross-section. (Recomputed 2026-09-28 after a compounding-convention audit; an earlier version of this box overstated our side by the annual-vs-continuous wedge and reported spurious 15-21 bp mid-curve gaps.)
Parameter sources: EIOPA, Report on the Calculation of the UFR for 2027 (30 March 2026); EIOPA RFR Technical Documentation EIOPA-BoS-25-599 (October 2025), Tables 2 and 7 and Section 9.4. Curve input: latest daily LW-NSS fit.
Source-by-source observation timestamps for every input feeding this page are listed in the Data lineage & freshness block on the landing page.